In recent months, securities class action litigation patterns involving AI-related disclosures have emerged and developed, as has been documented on this site (most recently, for example, here). There has of course been a great deal of other kinds of AI-related litigation, including lawsuits involving intellectual property issues, privacy and data-collection concerns, discrimination and bias claims, and a variety of different kinds of tort allegations.

In a shareholder derivative lawsuit recently filed against the board of software firm Adobe, these two lines of AI-related litigation crossed; the plaintiff shareholder alleges that the defendants violated their board duties by knowingly permitting the company to train its artificial intelligence tools using material copyrighted by others and in a way that subjected the company to IP-related litigation. As discussed below, the new lawsuit illustrates how the broader range of AI-related litigation can translate into follow-on D&O claims, representing yet another area of AI-related D&O risk.

Background

Adobe is a software firm that provides a variety of services including AI services. Among the firm’s AI services is its SlimLM series of small language model (SLM) software optimized for document assistance tasks on mobile devices. The firm’s SLM tools are trained, according to the subsequently filed securities lawsuit complaint, by “ingesting datasets that are chosen by Adobe personnel.” These training tasks involved the “collection and use of millions of lines of text.”

According to the complaint, among the datasets Adobe used to train its SLM tools are the Books3 dataset, which, according to the complaint, “contains hundreds of thousands of copyrighted books that were pirated without authorization or consent of the authors”; and the “Common Crawl” dataset, which, the complaint alleges “is also known to contain copyrighted and unauthorized materials.”

The complaint alleges that Adobe’s directors and officers “knew of the copyright violations being perpetrated through” Adobe’s SLM platform. Rather than “use a clean dataset,” Adobe’s executives, the complaint alleges, “followed a ‘ask forgiveness not approval’ model and adopted and implemented an unlawful plan to use a pirated data set” to develop the firm’s AI services. This misconduct, the complaint alleges, “constitutes bad faith and a violation of positive law” in “directing, approving, or permitting” the misconduct.

The complaint goes on to allege that this alleged misconduct has “resulted in multiple copyright holders filing lawsuits” against Adobe based on the firm’s alleged “failure to compensate [copyright holders] for downloading, copying, storing, or using their copyrighted works,” for which Adobe now faces “potential massive liability.” The company’s executives, the complaint alleges, were well aware of the issues with using the copyright holders’ copyrighted works, as well as of the potential liability for the company’s unauthorized use of copyrighted materials. The company has now been hit with multiple IP-related lawsuits filed on behalf of persons who allege that Adobe violated their copyrights.

The complaint alleges that after the first of several IP related lawsuits were filed against Adobe, its share price declined more than 25%. The complaint also alleges that as a result of “all of this misconduct and violations,” Adobe’s CEO and Chairman, Shantanu Narayen, stepped down from his position on March 12, 2026. Narayen’s departure was “specifically linked to Adobe’s failed AI strategy.”

The D&O Lawsuit

On April 24, 2026, a plaintiff shareholder filed a shareholder derivative lawsuit in the Northern District of California against Narayen and 13 other current and former officers and directors of Adobe, as well as against Adobe itself as nominal defendant. A copy of the complaint can be found here.

The complaint includes four substantive counts: first, a count alleging that all defendants violated their fiduciary duties; second, a count against all defendants alleging waste; third, a count against the director defendants alleging their violation of Section 14(a) of the Securities Exchange Act of 1934, in connection with the company’s proxy statements; and fourth, a count against defendants Narayen and Adobe CFO Daniel Durn for alleged violations of Section 10(b) of the Securities Exchange Act of 1934 and Rule 10b-5 thereunder in connection with alleged misrepresentations in the company’s disclosures and public statements.

The complaint seeks to recover damages Adobe allegedly sustained as a result of the alleged violations, and also seeks corporate governance reforms.

Discussion

The new Adobe derivative complaint has only just been filed and it remains to be seen how it will fare as it progresses. Readers of this site know well that derivative lawsuits are subject to formidable defenses, including, among other things, the demand requirement and the business judgment rule. The defendants undoubtedly will file motions to have the complaint dismissed based on these standard derivative suit defenses.

Regardless of how this complaint ultimately fares, the complaint itself is interesting for those whose jobs it is to worry about the potential liabilities of corporate directors and officers. That is, the complaint represents an example of how other kinds of AI-related liability exposures and litigation risks can translate into D&O claims.

It is not news that AI tool developers have been sued based on allegations that the companies have violated the copyrights of content owners by using the content to train their AI models. But while this kind of litigation is not new, what is new is for this type of litigation to lead to D&O related claims through follow-on litigation.

This development is important for what it represents in terms of potential AI-related D&O exposure. There is a great deal of discussion (and even arguably even agitation) in the insurance world in general, and in the D&O insurance world in particular, as industry professional try to assess and understand what kind of liability exposures will arise out of companies’ widespread adoption of AI strategies, AI processes, and AI tools. This new lawsuit, filed as a follow-on to intellectual property litigation involving AI-related allegations, represents a concrete example of a kind of D&O related liability exposure can arise from a company’s AI-based strategy.

It is important in thinking about the potential significance of this development to understand that the many other kinds of AI-related litigation – not only IP-related litigation, but privacy; discrimination and bias; and other types of tort litigation – could similarly lead to follow-on D&O litigation involving the same kinds of duty breach allegations raised in this complaint. To be sure, this phenomenon may be yet another example of the ways in which everything sooner or later becomes a D&O claim. While this observation about everything becoming a D&O claim may these days be a commonplace, that does not make it any less real.

The bottom line is that as allegedly aggrieved parties continue to conjure up ways in which they alleged that they have been harmed by artificial intelligence, these allegedly aggrieved parties’ claims may at least potentially trigger parallel follow-on D&O lawsuits, in which the underlying allegations are relied up on to support D&O liability claims.

As D&O professionals these days increasingly are called upon to assess how the widespread adoption of AI can contribute to D&O risk, it will be important to consider the risks associated with this type of AI-related follow-on litigation.